A long-standing goal in de novo protein design has been the exploitation of a hierarchical design strategy, utilizing appropriately designed peptide “building blocks,” to spontaneously assemble (via oligomerization or concatenation) the desired target architecture (DeGrado W. F et al. (1987). Cold Spring Harbor Symp. Quant. Biol. 52, 521-526, Regan, L. & DeGrado, W. F. (1988). Science, 241, 976-978, Richardson, J. & Richardson, D. C. (1989). Trends Biochem. Sci. 14, 304-309, Hecht, M. H. et al. (1990). Science, 249, 884-891, Quinn, T. P. et al. (1994). Proc. Natl Acad. Sci. USA, 91, 8747-8751, Bryson, J. W. et al. (1995). Science, 270, 935-941, Fu, X. et al. (2003). Protein Eng. 16, 971-977, Offredi, F. et al. (2003). J. Mol. Biol. 325, 163-174, Tsai, H. H. et al. (2004). Protein Sci. 13, 2753-2765, Heinemann, M. & Panke, S. (2006). Bioinformatics, 22, 2790-2799, Haspel, N. et al. (2007). Methods Mol. Biol. 350, 189-204, He, Y. et al. (2009). Comput. Biol. Chem. 33, 325-328 and Armstrong, C. T. et al. (2009). Faraday Discuss. 143, 305-317). Such hierarchical design strategies are “bottom-up” in that polypeptides are designed from first principles to have folding and thermodynamic properties that promote correct assembly of the target structure.
Symmetric or periodic protein architecture is favored for such de novo design and offers a number of potential advantages. A symmetric design constraint substantially reduces the conformational search in design algorithms and can simplify folding simulations, thereby substantially accelerating the design calculations (Fu, X. et al. (2003). Protein Eng. 16, 971-977, He, Y. et al. (2009). Comput. Biol. Chem. 33, 325-328 and Andre, I. et al. (2007). Proc. Natl. Acad. Sci. USA, 104, 17656-17661). Elements of symmetry that result in efficient structural compaction during folding likely contribute to an efficient funneled energy landscape of folding (Wolynes, P. G. (1996). Proc. Natl Acad. Sci. USA, 93, 14249-14255).
Structural symmetry can also result in multiple folding nuclei with an associated redundancy within the folding pathway (Lowe, A. R. & Itzhaki, L. S. (2007). Proc. Natl Acad. Sci. USA, 104, 2679-2684). However, there are also significant unresolved questions regarding the practical limitations of symmetric protein design. For example, exact primary-structure symmetry within a symmetric architecture involves a substantial reduction in sequence complexity—one hallmark of natively unstructured proteins (Wootton, J. C. & Federhen, S. (1993). Computers Chem. 17, 149-163 and Romero, P. et al. (2001). Proteins, 42, 38-48). Exact primary structure symmetry within repeated domains provides opportunities for domain mismatches producing misfolded forms with near-native Gibbs energy, and low sequence identities could have a crucial and general role in safeguarding proteins against misfolding and aggregation (Wright, C. F. et al. (2005). Nature, 438, 878-881); furthermore, primary-structure symmetry is one feature of amyloid-type aggregates (Hoang, T. X. et al. (2006). Proc. Natl Acad. Sci. USA, 103, 6883-6888). Thus, while symmetric protein architecture offers attractive advantages for de novo design, there is a need for novel approaches to successfully identify foldable peptide building blocks from those that might otherwise misfold or aggregate.
Citation or identification of any document in this application is not an admission that such document is available as prior art to the present invention.